Mean Reversion of Equity Prices: Less than Meets the Eye?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Mean Reversion of Equity Prices: Less than Meets the Eye?

Post by Browser »

The statistical evidence for the mean reversion of stock prices, upon which the assertion that stocks are less risky the longer they are held is based (Siegel: Stocks for the Long Run) and which is also put forth as one basis for "rebalancing" is not robust. It primarily depends on just a few historical "outliers" in the U.S. and U.K. in which large equity losses were eventually followed by large gains. However, in many other countries - the most recent notable example being Japan - large equity losses have never been followed by offsetting large gains, making the experience of the U.S. and U.K. outliers themselves.

Source: Credit Suisse Global Investment Returns Yearbook 2013
The US and UK stock markets have experienced
a few instances of dramatic reversals. In the USA,
there was a real capital loss of −67% (1929–32)
followed by a gain of +50% (1933). More recently,
there was a real capital loss of −39% (2008)
followed by a gain of +23% (2009). Similarly, in
the UK, there was a real capital loss of −36%
(1920) that was followed by a gain of +75%
(1921–22). And perhaps most dramatically, there
was Britain’s real capital loss of −74% (1973–74)
that was followed by a gain of +86% (1975).
We therefore check whether the mean reversion
we observe arises because of just
a very few brief historical episodes that may never
recur.
To a considerable extent, the in-sample pattern
of mean reversion in each of these markets is
thus attributable to just a couple of events per
market that occurred over the span of 113 years.

Moreover, collapses in these two markets were
followed by a recovery, and a relatively speedy
one at that. Investors in some other countries
were not so fortunate (think of China, Austria, or
perhaps Belgium). Evidently, the pattern of mean
reversal that we have uncovered is fragile. Even
on an in-sample basis, it depends critically on a
few outlying events.
We don't know where we are, or where we're going -- but we're making good time.
User avatar
Taylor Larimore
Posts: 32842
Joined: Tue Feb 27, 2007 7:09 pm
Location: Miami FL

John Bogle on "Reversion to the Mean"

Post by Taylor Larimore »

Browser:

This is John Bogle's opinion (in a 2002 speech I heard him give):

The Telltale Chart

Best wishes
Taylor
"Simplicity is the master key to financial success." -- Jack Bogle
Beagler
Posts: 3442
Joined: Sun Dec 21, 2008 6:39 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by Beagler »

Opinions vary.

http://tinyurl.com/a7w823t
"The Telltale Chart - http://www.vanguard.com/bogle_site/sp20020626.html

You should note that there are several problems with using the info in these charts to argue against tilting to small and value. One is that even Jack Bogle is not immune to cherry picking his data. You can see this most dramatically in his "RTM and "Slice and Dice" section, where he included gold as part of the S&D portfolio. Since the data ended in 2001, gold dragged the total return down. if you extended the same portfolio to today, you would find S&D to be a dramatic winner.

This suggests the even larger problem. The entire talk is really about Reversion to the Mean (RTM). He shows example after example of how different asset classes take turns outperforming and underperforming the total stock market (or the S&P 500, which is close to the same thing). Since they end up in the same place, he concludes you shouldn't bother going beyond TSM. But if you accept his premise that RTM is inevitable, you are also accepting that you will get greater returns without greater risk by holding multiple asset classes and rebalancing, since you always be selling (or buying less) of the currently overvalued asset and buying more (selling less) of the undervalued asset."


------------------------------------------------------------------------------------
Re: Request: John Bogle's speech/essay against Tilting (sm,

by Rick Ferri » Mon Jul 16, 2012 4:05 pm

I agree with Alex. I was at the Morningstar conference when Jack gave the speech and had to chuckle at all the data-mining Jack went through to come up with a slide to support his conclusion.

Rick Ferri
“The only place where success come before work is in the dictionary.” Abraham Lincoln. This post does not provide advice for specific individual situations and should not be construed as doing so.
Topic Author
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by Browser »

Of course one of the problems with RTM is determining just what "mean" stock prices are reverting to. When we see the chart of stock returns, for example, from 1926 to 2012, with hindsight we can fit a trend line to the returns. A compelling interpretation is that there is some sort of "gravitational pull" that kept bringing prices back to the trendline. But if you were to look at a chart of stock returns over a different periods of time, you would see many different trendlines. All we in fact know is that stock returns vary year by year, and in the USA and UK at least they have tended to go up over time. Whether this follows any kind of systematic mean reversion pattern is in the eye of the beholder.

What is being ignored is that the trendline that is being used to "prove" mean reversion is one that we couldn't have predicted ex ante. There was absolutely no way to know, in 1926 for example, whether USA stock prices would go up, down, sideways, in a corkscrew pattern, a cliff-diving pattern or any other; and there was certainly no way to know where the price would be at any point in the future. So "mean reversion" is a pattern that only exists ex post, only in hindsight. It has absolutely no predictive value.

Another important shortcoming of the "research" showing mean reversion is that it is invariably based on using charts of U.S stocks or U.K. stocks rather than using global stock returns. Inferences made on the basis of those charts alone runs the risk of being spurious. An economist looking instead at the historical stock price charts of Austria, Russia, China, and many other world equity markets including Japan during the 20th century would see stock prices that collapsed and never recovered. That economist would likely not have formulated a "mean reversion" hypothesis to explain stock price movements.
We don't know where we are, or where we're going -- but we're making good time.
User avatar
nisiprius
Advisory Board
Posts: 52216
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by nisiprius »

I'm delighted by Browser's post. I've been skeptical all along. Just as people have been talking about "momentum," by that word, in securities prices at least since the early 1900s, "mean reversion" is not a new concept either. It wasn't called by that name, but Charles Babson had an admiration for Sir Isaac Newton and believed that Newton's Third Law--for every action, there is an equal and opposite reaction--applied to the stock market. It's hard to believe that these effects, if they really existed, aren't so well known that they would long have become common knowledge and vanished.
Beagler wrote:You can see this most dramatically in his "RTM and "Slice and Dice" section, where he included gold as part of the S&D portfolio. Since the data ended in 2001, gold dragged the total return down. if you extended the same portfolio to today, you would find S&D to be a dramatic winner.
You don't have to make it hypothetical. What does John C. Bogle enormous credit, in my view, is that in his 2011 book, Don't Count On It!, Bogle reprinted "The Telltale Chart"--with all of the data updated to 2009. And as Beagler says, S&D outperformed. And he includes an update within his text, where he acknowledges, explicitly, "Through 2009, the soaring price of gold has helped offset the below-average returns of the alternative slice-and-dice portfolio."

Image
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
grayfox
Posts: 5569
Joined: Sat Sep 15, 2007 4:30 am

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by grayfox »

My understanding is that one model for stock returns is a random walk model.

Now for a random walk, the variance of wealth grows linearly with time. But people who look at this have observed that variance grows less than linearly with time. They do some kind of Variance-Ratio test. So they postulate that there must be some kind of mean reversion going on.

Exactly how the mean reversion works, I'm not sure.
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by larryswedroe »

First, the idea that stocks are less risky the longer the horizon is simply wrong. It relies on the idea that SD is lower over time for stocks. But that is irrelevant, the longer the horizon the wider the dispersion of potential outcomes, which of course makes stocks riskier.

Second RTM is LIKELY in stocks for the simple reason that valuations matter. When returns are above average that likely means valuations rose and that means future expected returns are now LOWER, and vice versa. But while higher past returns does mean lower future returns (assuming it's valuation driven) low returns don't guarantee high future returns, markets can implode and go to zero for example.

Best wishes
Larry
User avatar
jidina80
Posts: 729
Joined: Wed Feb 24, 2010 4:05 pm
Location: Fiji

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by jidina80 »

larryswedroe wrote:First, the idea that stocks are less risky the longer the horizon is simply wrong. It relies on the idea that SD is lower over time for stocks. But that is irrelevant, the longer the horizon the wider the dispersion of potential outcomes, which of course makes stocks riskier.
Stocks do have more variability in returns than bonds, but almost all the long term dispersion of outcomes is better than bonds. Over the long term, one has more risk in not achieving financial security by investing only in bonds.

Who thinks stocks are riskier in the long term when the wide dispersion of potential outcomes are almost always better than bonds?
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by larryswedroe »

jidina
That is simply the triumph of the optimists story, more risk is EXPECTED to have higher returns, better outcomes or people would not invest. That doesn't change the fact that it's more risky.
Now that is a different story if you ask if I choose stocks or bonds what are my odds of achieving a certain level of wealth? Of course stocks give you a greater chance of achieving a higher level of wealth. But they also give you a higher chance of having a much lower level.
Larry
User avatar
tadamsmar
Posts: 9972
Joined: Mon May 07, 2007 12:33 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by tadamsmar »

Browser wrote:Of course one of the problems with RTM is determining just what "mean" stock prices are reverting to. When we see the chart of stock returns, for example, from 1926 to 2012, with hindsight we can fit a trend line to the returns. A compelling interpretation is that there is some sort of "gravitational pull" that kept bringing prices back to the trendline.
If this is what RTM means, then its a mathematical certainty. Consider any sequence like 3,1,4,5,9. Calculate the mean. If the sequence is a time series, then you will see that RTM is an elementary fact of arithmetic.

Does not seem to me that this tautology is of any particular importance to investors
User avatar
tadamsmar
Posts: 9972
Joined: Mon May 07, 2007 12:33 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by tadamsmar »

tadamsmar wrote:
Browser wrote:Of course one of the problems with RTM is determining just what "mean" stock prices are reverting to. When we see the chart of stock returns, for example, from 1926 to 2012, with hindsight we can fit a trend line to the returns. A compelling interpretation is that there is some sort of "gravitational pull" that kept bringing prices back to the trendline.
If this is what RTM means, then its a mathematical certainty. Consider any sequence like 3,1,4,5,9. Calculate the mean. If the sequence is a time series, then you will see that RTM is an elementary fact of arithmetic.

Does not seem to me that this tautology is of any particular importance to investors
Heck, if you add up that series of 5 returns and you determine the growth over the 5 time periods (5 times the mean), then you see that the total returns will always return exactly to the mean for the total period. This is just obvious math.
User avatar
tadamsmar
Posts: 9972
Joined: Mon May 07, 2007 12:33 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by tadamsmar »

There is another meaning for RTM. Bogle is using it here:

"Of course, mutual fund marketers assume—partially correctly—that most investors are completely unaware that today's top performers are overwhelmingly likely both to be tomorrow's ordinary participants in the stock market, and to parallel the average of their peers—in other words, that today's Beau Brummels are tomorrow's Joe Six-Packs."

http://www.vanguard.com/bogle_site/lib/sp19980129.html

This expresses the notion that past returns don't predict future returns. The good and bad performers of the past all revert to the average future performance of their peers.

This is not based on using some past mean as a predictor. It says forget about the past relative means as a tool to predict the future relative means.

Random walks evidence this type of RTM.
User avatar
tadamsmar
Posts: 9972
Joined: Mon May 07, 2007 12:33 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by tadamsmar »

grayfox wrote:My understanding is that one model for stock returns is a random walk model.

Now for a random walk, the variance of wealth grows linearly with time. But people who look at this have observed that variance grows less than linearly with time. They do some kind of Variance-Ratio test. So they postulate that there must be some kind of mean reversion going on.

Exactly how the mean reversion works, I'm not sure.
The evidence for rejecting random walk from variance-ratio tests is not conclusive, according to this paper:

http://www.kansascityfed.org/PUBLICAT/E ... 91enge.pdf
User avatar
tadamsmar
Posts: 9972
Joined: Mon May 07, 2007 12:33 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by tadamsmar »

I just posted about 3 different meanings for RTM. Note that Bogle seems to have multiple things in mind when he uses the term:
RTM represents the operation of a kind of "law of gravity" in the stock market, through which returns mysteriously seem to be drawn to norms of one kind or another over time.
http://www.vanguard.com/bogle_site/lib/sp19980129.html

Interestingly, the various meanings are contradictory. Some support the idea that the market is a random walk, some support the idea that you should time investments based on "valuations". Others are just obvious mathematical facts. Can all of these really be kind of like laws of gravity?
Calm Man
Posts: 2917
Joined: Wed Sep 19, 2012 9:35 am

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by Calm Man »

Rather than calling it mean reversion, couldn't we simply view it as stock prices eventually returning to their "proper" relationship to earnings? This I believe is Buffet's idea that he doesn't care if the stock market goes to sleep for 10 years. So 50 years from now, whether in part from inflation or real gains, if earnings are higher, then stock prices should be higher. No?
User avatar
tadamsmar
Posts: 9972
Joined: Mon May 07, 2007 12:33 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by tadamsmar »

Calm Man wrote:Rather than calling it mean reversion, couldn't we simply view it as stock prices eventually returning to their "proper" relationship to earnings? This I believe is Buffet's idea that he doesn't care if the stock market goes to sleep for 10 years. So 50 years from now, whether in part from inflation or real gains, if earnings are higher, then stock prices should be higher. No?
You tell me. BP's earnings were not immediately impacted by the Deepwater Horizon disaster, but it's price sure was. Should their be a role for projected earnings and risks in stock price setting?
JimInIllinois
Posts: 405
Joined: Wed Dec 22, 2010 8:55 am

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by JimInIllinois »

tadamsmar wrote:The good and bad performers of the past all revert to the average future performance of their peers.

This is not based on using some past mean as a predictor. It say forget about the past relative means as a tool to predict the future relative means.

Random walks evidence this type of RTM.
Exactly. Mean-reversion only means that exceptional performance is likely to be followed by average performance, not that future performance will compensate for past performance. The hard part is predicting what average performance will be in the future.

One asset that is mean-reverting in the strongest possible sense is a zero-coupon treasury bond, since its value at maturity is known exactly and average future return can be trivially calculated.

The argument for stronger mean-reversion in stocks is that earnings and book values are much more stable than prices, and that the multiples of prices relative to expected earnings and book value are constrained by the self-interested behavior of market participants. The true long-term risks are total economic collapse and government confiscation, both of which were real fears during the great depression.
Topic Author
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by Browser »

tadamsmar wrote:
tadamsmar wrote:
Browser wrote:Of course one of the problems with RTM is determining just what "mean" stock prices are reverting to. When we see the chart of stock returns, for example, from 1926 to 2012, with hindsight we can fit a trend line to the returns. A compelling interpretation is that there is some sort of "gravitational pull" that kept bringing prices back to the trendline.
If this is what RTM means, then its a mathematical certainty. Consider any sequence like 3,1,4,5,9. Calculate the mean. If the sequence is a time series, then you will see that RTM is an elementary fact of arithmetic.

Does not seem to me that this tautology is of any particular importance to investors
Heck, if you add up that series of 5 returns and you determine the growth over the 5 time periods (5 times the mean), then you see that the total returns will always return exactly to the mean for the total period. This is just obvious math.
Yes, you are exactly correct. Mean reversion in the mathematical sense is trivial. And it can only be demonstrated ex post, once we know the beginning point and the end point of stock returns.
We don't know where we are, or where we're going -- but we're making good time.
JimInIllinois
Posts: 405
Joined: Wed Dec 22, 2010 8:55 am

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by JimInIllinois »

tadamsmar wrote:Should their be a role for projected earnings and risks in stock price setting?
Of course. The only role that past earnings play is to help predict future earnings.
Topic Author
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by Browser »

[Nikkei 225 Stock Index - 22 Year Graph (Japan)
Image

My financial advisor told me that prices would revert to the mean -- but he didn't say which mean. :oops:
We don't know where we are, or where we're going -- but we're making good time.
dkturner
Posts: 1940
Joined: Sun Feb 25, 2007 6:58 pm

Re: John Bogle on "Reversion to the Mean"

Post by dkturner »

Taylor Larimore wrote:Browser:

This is John Bogle's opinion (in a 2002 speech I heard him give):

The Telltale Chart

Best wishes
Taylor
Taylor, is it ever possible to have an intelligent discussion about a financial research paper without someone dragging Jack Bogle's opinion (not research) into it?
User avatar
vectorizer
Posts: 512
Joined: Sat Mar 03, 2007 2:52 pm

Re: John Bogle on "Reversion to the Mean"

Post by vectorizer »

dkturner wrote:Taylor, is it ever possible to have an intelligent discussion about a financial research paper without someone dragging Jack Bogle's opinion (not research) into it?
Since this is the bogleheads board, I hope the answer is "no".
Calm Man
Posts: 2917
Joined: Wed Sep 19, 2012 9:35 am

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by Calm Man »

tadamsmar wrote:
Calm Man wrote:Rather than calling it mean reversion, couldn't we simply view it as stock prices eventually returning to their "proper" relationship to earnings? This I believe is Buffet's idea that he doesn't care if the stock market goes to sleep for 10 years. So 50 years from now, whether in part from inflation or real gains, if earnings are higher, then stock prices should be higher. No?
You tell me. BP's earnings were not immediately impacted by the Deepwater Horizon disaster, but it's price sure was. Should their be a role for projected earnings and risks in stock price setting?
I haven't followed BP. I know it went down a lot right away. I figure it probably has recovered some. I am referring more to the market as a whole. Say S&P 600 or total stock market earnings. That is what most of us invest in (TSM). Basically don't prices ultimately reflect earnings times some multiple? The multiple might be from 10-15 or so. But won't the index earnings be higher almost certainly in 30 years?
User avatar
Taylor Larimore
Posts: 32842
Joined: Tue Feb 27, 2007 7:09 pm
Location: Miami FL

Re: John Bogle on "Reversion to the Mean"

Post by Taylor Larimore »

Taylor, is it ever possible to have an intelligent discussion about a financial research paper without someone dragging Jack Bogle's opinion (not research) into it?
dk turner:

It is unfortunate that our mentor's opinion on the topic offends you.

Best wishes.
Taylor
"Simplicity is the master key to financial success." -- Jack Bogle
Topic Author
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by Browser »

Thanks Taylor -

Can't speak for everyone, but I'd surmise that very few of us are offended by anything Jack Bogle has to say. :beer
We don't know where we are, or where we're going -- but we're making good time.
garlandwhizzer
Posts: 3565
Joined: Fri Aug 06, 2010 3:42 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by garlandwhizzer »

P/E ratios and other good measures of fundamental valuation do in fact tend to revert to the mean. The Japan example of 1990 to present that everyone likes to use to deny mean reversion was in fact a mean reversion in P/E ratios from outrageous P/E levels (>60) to present levels that are I believe lower than the US. Hyper-inflated market valuations like Japan in 1989 or US tech stocks in 1999 or the housing market in 2006 take a long time to revert to the mean, but at the inflection point the force of gravity always prevails over optimistic sentiment tugging down on excessive valuations. Depending on how severely inefficient the market was and the macroeconomic status of growth going forward, it may take a decade or two or three or more to correct.

The simple fact is that valuations matter and when they depart enormously from historical means in either direction, the prevailing trend is likely to shift powerfully in the opposite direction.

Garland Whizzer
Topic Author
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by Browser »

Better not hope for mean reversion if you're looking at the mean trend from 1958 to 1994. Lot of mean reverting left to do, and you'll be mean reverted into poverty!

Image
We don't know where we are, or where we're going -- but we're making good time.
User avatar
Taylor Larimore
Posts: 32842
Joined: Tue Feb 27, 2007 7:09 pm
Location: Miami FL

Different charts. Different picture.

Post by Taylor Larimore »

Bogleheads:

Browser's chart is linear.

Jeremy Siegel's chart is logarithmic (page 4).

http://www.cobank.com/~/media/Files/Sea ... Siegel.pdf

Best wishes
Taylor
"Simplicity is the master key to financial success." -- Jack Bogle
garlandwhizzer
Posts: 3565
Joined: Fri Aug 06, 2010 3:42 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by garlandwhizzer »

Fascinating post, Taylor. Thanks.

Jeremy Siegel is a knowledgeable cheerleader for stocks and particularly high dividend stocks and value stocks. His long term charts in this post tend to substantiate his point of view at least over very long periods of time. Also his concern about the future of developed economies in the face of retirement burdens and population demographics is in my opinion right on target. I suspect that only a small percentage of my generation, the baby boomers, are in a realistic financial position to retire without some probability of running out to money. The same with pension funds, which as the Credit Swisse Global Investment Returns Yearbook 2013 points out, are woefully underfunded and have unrealistic assumptions for future returns. The sky is not falling, but there are plenty of challenges going forward. Portfolios that have worked reasonably well in the past decade or so, may not be counted on to do so well going forward.

We live in interesting times undoubtedly. It is good that we have solid real time informational and insight sources like this forum to help us make wise financial decisions in order to cope with the uncertainty that the future holds.

Garland Whizzer
User avatar
tadamsmar
Posts: 9972
Joined: Mon May 07, 2007 12:33 pm

Re: Different charts. Different picture.

Post by tadamsmar »

Taylor Larimore wrote:Bogleheads:

Browser's chart is linear.

Jeremy Siegel's chart is logarithmic (page 4).

http://www.cobank.com/~/media/Files/Sea ... Siegel.pdf

Best wishes
Taylor
The squiggly while line is the stock growth line according to the caption. One gets the impression that the red straight line over it is the trend line for the squiggly line, but it can't be because it begins at or below the beginning of the white line and ends above it. As I pointed out earlier, the trend line for any squiggly line ends and begins exactly on the squiggly line no matter if the squiggly line is a stock time series or just any squiggly line that somebody scrawled, by virtue of simple mathematical rules. If the red line is a trend line, then it represents a trend line for something that grew faster than the white stock line over the plotted period.
User avatar
Ketawa
Posts: 2521
Joined: Mon Aug 22, 2011 1:11 am
Location: DC

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by Ketawa »

Couldn't it just be a MMSE linear regression line? Then it doesn't need to start/end on the data series.
Topic Author
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by Browser »

From the linked article:
Conditional on knowing the trend rate of return,
“forecasts” based on whether stocks are deemed
“cheap” or “expensive” will be completely accurate.
By construction, equity prices will at a future
date revert to the long-term mean. While we do
not know the speed of mean reversion, we know it
must happen by the end-date of the long-term
return series. However, as an investment system,
this approach is inoperable as it requires the investor
to be prescient about the eventual performance
of the stock market. The temptation to fit
such trendlines seems irresistible. Unfortunately,
they mislead, rather than inform.
We are tempted to recall the comment: "forecasters who depend on historical data to prove their predictions are using statistics like drunks use lampposts - for support rather than illumination."
We don't know where we are, or where we're going -- but we're making good time.
ourbrooks
Posts: 1575
Joined: Fri Nov 13, 2009 3:56 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by ourbrooks »

Let's see, we can't depend on stocks because mean reversion might not occur and stocks might just go steadily downward.

From the same document, read their analysis on bonds, particularly pages 6 and 7 of the first article. They claim that the performance of bonds over the last 30 years or so is a historical anomaly, distorted by the "race to zero" of interests rates over the past dozen years. Their projections are that short bonds will have negative real returns over the next 6-8 years and that even over the next 20 years, real returns are likely to be less than 1%. That's for long bonds; for rolling short bonds, the expected return is negative.

If any of these analyses are to be believed, an investor's choices these days are either guaranteed slow impoverishment from bonds or a chance at either modest wealth or rapid poverty from stocks. Hmm, maybe I was a bit hasty in dismissing gold as a holding.
Topic Author
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by Browser »

Just for the record, Siegel's data is not exactly above reproach, as documented in an article by Jason Zweig

Does Stock-Market Data Really Go Back 200 Years?
There is just one problem with tracing stock performance all the way back to 1802: It isn't really valid.

Prof. Siegel based his early numbers on data first gathered decades ago by two economists, Walter Buckingham Smith and Arthur Harrison Cole.
The Smith and Cole indexes are [not comprehensive or representative], as the professors signaled in their 1935 book, "Fluctuations in American Business." They cherry-picked their indexes by throwing out any stock that didn't survive for the whole period, whose share prices were too hard to find or whose returns seemed "inflexible," "erratic," or "non-typical."

Here is a second problem with Prof. Siegel's data.
In an article published in 1992, he estimated the average annual dividend yield from 1802-1870 at 5.0%. Two years later in his book, it had grown to 6.4% -- raising the average annual return in the early years from 5.7% to 7.0% after inflation.

The 1802-to-1870 stock indexes are rotten with methodological flaws. So we have only the period since then, or four distinct and complete 30-year stretches of stock returns, to base our long-term investment decisions on.

Another emperor of the late bull market, it seems, has turned out to have no clothes
We don't know where we are, or where we're going -- but we're making good time.
adam1712
Posts: 566
Joined: Fri Jun 01, 2007 5:21 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by adam1712 »

larryswedroe wrote:jidina
That is simply the triumph of the optimists story, more risk is EXPECTED to have higher returns, better outcomes or people would not invest. That doesn't change the fact that it's more risky.
Now that is a different story if you ask if I choose stocks or bonds what are my odds of achieving a certain level of wealth? Of course stocks give you a greater chance of achieving a higher level of wealth. But they also give you a higher chance of having a much lower level.
Larry
But the probability that stocks will have a worse return than bonds over a 30 year period is less than the probability of one year stock returns being worse than bonds. It's not just that expected average returns are higher, it's that the probability of stocks being a worse choice than bonds decreases with time. This is with a simple random walk model, no reversion to the mean necessary.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by Rodc »

I confess I'm just skimming.

One way you can "see" mean reversion by looking at correlations of rolling independent periods. If you do this you start to see small but meaningful negative correlation out around 20 years. A good 20 year period tends (slightly) to be followed by a not so good 20 year period and vice versa.

The problem is we only have about 4 or 5 decent independent 20-year data points. Now some very modest mean reversion forcing function may be at work (20 years is long enough for those scared by bad times to get replaced by a new crop of stock enthusiasts who drive up prices, eventually get clobbered, and it takes 20 years to grow a new crop of stock enthusiasts, :) ) But you can't prove anything, or even make a strong case, with 4 or 5 data points.

I also note that if you have a scattering of data points and you draw a line through them it is very easy to get a picture that looks like it has mean reversion forcing function at work. Make some random values and give it a try.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
lostcowboy
Posts: 213
Joined: Tue Sep 16, 2008 1:30 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by lostcowboy »

I like how they use the word "real" instead of "raw" prices. Maybe the raw prices have more mean reversion to them? Remember, in the market place you can only deal with the raw prices, you don't get to deal with the real prices.
Akiva
Posts: 536
Joined: Tue Feb 15, 2011 1:33 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by Akiva »

Browser wrote:The statistical evidence for the mean reversion of stock prices, upon which the assertion that stocks are less risky the longer they are held is based (Siegel: Stocks for the Long Run) and which is also put forth as one basis for "rebalancing" is not robust. It primarily depends on just a few historical "outliers" in the U.S. and U.K. in which large equity losses were eventually followed by large gains. However, in many other countries - the most recent notable example being Japan - large equity losses have never been followed by offsetting large gains, making the experience of the U.S. and U.K. outliers themselves.
Regression to the mean is a poorly understood phenomena, and the "mean reversion" that book is talking about isn't even the real thing. Siegel basically invented a new meaning for the term and used it because it suited his argument. It isn't surprising that a made up phenomena isn't actually supported by the data. Anyone familiar with the actual data wouldn't have needed this article to tell you that the random walk model is a better fit than Siegel's "mean reversion" one.
Topic Author
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Mean Reversion of Equity Prices: Less than Meets the Ey

Post by Browser »

There are many myths about stock investing, and I think RTM is one of them. I figure most of what we do as investors is based on untruths and downright lies that we believe. But that's probably a good thing, otherwise we'd realize it's all probably a crapshoot and we wouldn't be able to act at all.
We don't know where we are, or where we're going -- but we're making good time.
Post Reply